Sensorimotor learning allows organisms to interact adaptively with their environment, yet a detailed description of how sensory input modifies motor circuitry during learning has not been developed. Since many movement disorders involve dysfunction of motor planning and disruption of learned motor programs, understanding how motor programs are instantiated in motor circuits during learning is essential. Vocal learning in songbirds is a particularly tractable model of sensorimotor learning since birdsong is an ethological and readily quantifiable behavior, and because the neuroanatomical substrates of song production are identified and accessible. Song learning entails the convergence of juvenile song toward a tutor song, and although changes in song are associated with changes in premotor nuclei, the synaptic and cellular plasticity mechanisms underlying song learning remain unknown. This proposal describes experiments that will combine a form of rapid song learning with chronic in vivo electrophysiology in behaving animals to identify patterns of premotor activity that accompany changes in song, and utilize in vitro whole-cell recording and minimal stimulation to identify the synaptic loci and mechanisms of this plasticity. Results from these experiments will provide an integrated and comprehensive account of vocal motor learning, from synaptic plasticity to changes in circuit dynamics to altered vocal behavior, and thereby give broad insight into how the nervous system learns adaptive and complex behaviors.